Characterization of DNA sequence-common and sequence-specific proteins binding to cis-acting sites for cleavage of the terminal a sequence of the herpes simplex virus 1 genome.

0022-538XI891031059-10$02.00/0

CopyrightC 1989, AmericanSocietyforMicrobiology

Characterization

of DNA

Sequence-Common

and

Sequence-Specific

Proteins Binding

to

cis-Acting

Sites

for

Cleavage of the Terminal a

Sequence of

the Herpes

Simplex Virus

1

Genome

JOANY CHOUAND BERNARDROIZMAN*

The

Marjorie

B. Kovler ViralOncologyLaboratories, the Universityof Chicago, 910 East58th Street,

Chicago,

Illinois

60637

Received13 September 1988/Accepted 12 November 1988

The terminal 500-base-pair a sequence of the herpes simplex virus 1 genome contains signals for cleavage (Pacl and Pac2) of unit-length DNA molecules from concatemers inuniquestretches ofsequencesdesignated Ub and

Uc,

respectively, and a cis site forcleavage designated DR1. We report that nuclear extracts from infected cellscontain factors which form two DNA-virus-specific protein complexes with components of the a sequence. Purffication of the factors forming the V2 complex yielded a protein with an apparent molecular weight of 82,000 binding to DNA in a non-sequence-specific manner. Addition of

Mg2+

to the purified protein-DNA probe mixture resulted inexonucleolytic degradation of the DNA. The protein was identified as thevirus-specific DNase with monoclonal antibodyspecific for the viral enzyme. The purification of the proteins forming the V4 complex yielded two proteins with molecular weights of >250,000 and 140,000 corresponding to infected cellprotein 1 and to an as yet unidentified protein, respectively. These proteins formed two DNA sequence-common bands with a number of DNA probes and onesequence-specific band with probes containing both Pac2 andDR1 but not withprobes containing either site alone or Pacl and DR1. Since the DNA probe containing Pac2 and DR1 inserted into viral genome or into amplicons induced specific cleavage of the DR1 sequence whereas the nonreactive probes failed to induce thecleavage, the formation of this sequence-specific DNA-protein complex is significant and may reflect a DNA-protein interaction essential for cleavage. The possible role of the proteins identified in this study for the cleavage-packaging of viral DNA into capsids is presented.

The terminal sequence, designated sequence a, of the

herpes simplex virus 1 (HSV-1) genome (Fig. 1) encodes

several cis-acting sites involved in (i) the circularization of

viral DNA after infection, (ii) the inversion of the two

covalently linkedcomponentsofHSV-1DNA, long(L) and

short(S), relativetoeachother,(iii)thecleavage of

genom-ic-lengthDNAfromcirclesorconcatemersandpackaging of

theexcised molecules into capsids, and(iv)theexpression of

anmRNAextending from theasequenceintothereiterated

sequenceofthe L component(2-7, 21-23, 25, 26, 28,29). In

the expectation that specific viral or host proteins or both

interact with the DNA at the specific cis sites, we

con-structedseveral viralDNAprobes

encompassing

mostofthe a sequence and tested these for

reactivity

with proteins

contained innuclear extractsofmock-infected and infected

cells.Wereport that twoDNA-protein complexes formed by

the probes with nuclear extracts of infected cells each

contain a

virus-specified

protein. The

properties

of these

proteinsare described.

Relevant to this report are the structure and nucleotide

sequence arrangementofthe a sequence. InHSV-1 strain F

[HSV-1(F)],

theasequencesituatedatthe

junction

between the L and S components andrepeated inan inverted

orien-tationatthetermini ofthe genomeconsistsofa

20-base-pair

(bp) terminal repeat

designated

DR1, a

64-bp unique

se-quence designated Ub, 22 repeats of a 12-bp sequence

designated

direct repeat 2 (DR2), three repeats ofa

37-bp

sequence

designated

directrepeat 4(DR4)and

containing

11 of the 12 nucleotides of DR2, a

unique

58-bp

sequence

*Corresponding author.

designated

Uc,

and a second copy of DR1 (23). Tandem

repeats ofthe a sequence share the intervening DR1. The number ofcopies ofDR2 and DR4 vary, andpartial copiesof these sequences may also be repeated. This variability accountsformostoftheobserved a sequencepolymorphism (4-6, 21, 23, 28).

The observations that theterminal DR1 ofthe L

compo-nent consists of18 bp plus a single nucleotide extended 3'

andthatthe terminalDR1oftheScomponentconsists of1

bp plus a single nucleotide extended 3' producing, when

ligated,anintactDR1 led tothe conclusionthat the

genomic-lengthDNAisexcised fromconcatemersbycleavage within

DR1 sequences(23). Recentstudieshavedemonstratedthat

the signals for the cleavage and

packaging

of the DNA

designatedPac2 andPacl(Fig. 1)arelocatedin the

Uc

and

Ubsequences, respectively, of thea sequence(6,28).

(These resultswerepresentedattheInternational

Herpes-virusWorkshop, Irvine, Calif., August 1988.) MATERIALS ANDMETHODS

Cells, viruses, and protein extracts. HeLa cells grown to

confluency in 850-cm2 roller bottles were either mock

in-fected or inin-fected with 5 PFU ofHSV-1(F)orherpes

simplex

virus (HSV-2) strain G

[HSV-2(G)]

per cell. Cells were

harvested 20 to 24 h afterinfection ormock

infection,

and nuclear extracts werepreparedin 0.42 MKCI asdescribed previously (8).

Preparation ofDNAprobesandcompetitorDNAand DNA band shift assays for DNA-binding proteins. DNA

probes

used in thetext wereprepared

by

digesting

the

plasmids

with 1059

on November 10, 2019 by guest

http://jvi.asm.org/

1060 CHOU AND ROIZMAN

F probe

z-a,omb

L ">

S@

C

pac-I A probe pac-2

DRI

U

(DR2)22

(DR4)

DRI

X Dral

ApaI

__ .I

Eprobe pac-2 1

H

probe

Hprobe rr DR4

DRIDRI

G probe

[1

BamHI

AvaUI

ApaI

BamHI

Jprobe C probe

D probe

B probe Kprobe

l

L probe

A

FIG. 1. Schematic diagram ofthe sequence arrangementsin the HSV-1genomeand in theasequenceand thepositionin thegenomeof the DNAprobes used in this study. Top line:Sequencearrangementof the HSV-1genome.a,anda.refertotheterminalasequencesofthe LandScomponents.Subscriptn, One or moreasequences;subscriptm,noneto morethanone asequences.The insert indicates theposition ofthe F probe consisting of the 48 bp ofthe regulatory domain ofthe HSV-1(F) a27gene frompRB606 (13). Second line: Sequence arrangementof theasequence.Thewild-typeasequencecontainsa20-bp directly repeatedelement(DR1)flanking theasequence;a64-bp uniquesequence(Ub); 22copies of the12-bpdirectlyrepeatedelement(DR2);3copies of the 12-bp directly repeated element (DR4); anda

58-bpuniquesequence(Uc). The AprobeencompassingthreecopiesofDR4and onecopyofDR2wasderived fromtheplasmidA-1 and generated by BssHII andApalrestrictiondigestoftheA-1 clone(6).The DandEprobesderived frompRB3389andpRB3387,respectively, weregenerated by cleaving the parentplasmid withBamHI forprobe DandBamHI toApaIfor probe E.The pRBplasmids have been previously described (2). TheLprobewasgenerated byBamHIandApal cleavageofaplasmidwhich containsaBssHII deletion ofpRB3389 as shown; the C, J, B, and K probes were generated from pRB3389 by taking theBamHI-to-AvaII subfragment, the AvaII-to-BamHI fragment, the BamHI-to-ApaIfragment, and theApaI-to-BamHIfragment, respectively, asindicated. TheGand Hprobesweregenerated from pRB3387 bytaking the BamHI-to-DraI fragmentandtheDral-to-Apal fragment, respectively.

appropriate restriction enzymes. The DNAs were then

de-phosphorylated with calf intestinal alkaline phosphatase

(Boehringer Mannheim Biochemicals, Indianapolis, Ind.) andloadedona5% nondenaturingpolyacrylamide gel.The

fragmentsof interestwerethenpurified frompolyacrylamide

gels and 5' end labeled with [32P]ATP (>7,000 Ci/mmol; Dupont, NEN Research Products, Boston, Mass.) and T4

polynucleotide kinase (United States Biochemical Corp.,

Cleveland, Ohio) to an activity of approximately 30,000

cpm/ngof DNAfragment(17).

The standard competitor DNA usedin these assays was

synthetic poly(dI-dC)- poly(dI-dC) (Pharmacia, Molecular Biology Division, Piscataway, N.J.). Competitor polymers

were used at 2 to 3 ,ug per reaction of standard nuclear

extract and 0.5 ,ug per reaction ofpurified proteins in gel binding retardationassays.

Binding reaction mixtures (12pl)contained approximately 0.1ngoflabeledDNA, 2to3,ug of competitor DNA, 9

RI

of binding reaction buffer (20 mM Trishydrochloride[pH 7.4], 50 mM KCl, 0.05% Nonidet P-40, 5% glycerol, 50 ,ug of bovine serum albuminper ml, 5 mM 3-mercaptoethanol, 1 mMEDTA), and 2to 3 ,ug ofextractprotein. After incuba-tionatroomtemperaturefor 30min,reaction mixtureswere

loadedon a5% polyacrylamide gel in 50 mM Tris-borate-1 mMEDTAbuffer(pH 8.3)prerunat230 V for 2 h. Gelswere

run at 230 V until bromophenol blue dye reached the gel

bottom; then they were dried and autoradiographed (27).

Whenmonoclonal

antibody

wasusedin the

binding reaction,

1

RI

ofspecificdilutedmonoclonalantibodywasadded tothe

reactionmixture aftertheincubation periodin the standard

reactions and the reaction was allowed to incubate for

additional30minbefore

loading

ontogels.

Chromatographic purificationofproteins. The

purification

scheme used in these studies is shown in

Fig.

2. For P11

phosphocellulose chromatography, 50 to 70 mg of nuclear

extract stored at 12.9 mg/ml in buffer D (20 mM Tris

hydrochloride

[pH 7.9],

0.42M

KCl,

1.5mM

MgCI2,

0.2mM

EDTA, 0.5 mMdithiothreitol,0.5mM

phenylmethylsulfonyl

fluoride,25%glycerol)wasdiluted with buffer C (20mMTris

hydrochloride

[pH

7.9],

0.2 mMEDTA,0.5 mM

dithiothre-itol, 25% glycerol,0.5 mMphenylmethylsulfonylfluoride)to

afinalKCIconcentration of0.125 M. Thedilutedextractwas

loadedontoapreequilibrated P11 phosphocellulose column

(Whatman, Inc., Clifton, N.J.) prepared according to the

instructions ofthe manufacturer. The column was washed

with 2 column volumesofbuffer Ccontaining0.15 M

KCl.

A

stepgradientinbuffer C

containing

KCl

ranging

from0.15to

0.6 Min stepsof0.05Mwasthen

applied,

and the fractions

were collected. The fractions were then analyzed for the

J.VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:2.612.152.480.71.349.2]

HSV(1) Nuclear

Extract

Pi1 buffer C

0.15M KC1

Flowthrough 0.25-0.30M KC1

buffer c Hepa ri n buffer C

0.15M KC1 Agarose 0.15M KC1

0.30-0.35M KC1 0.35-0.40M KC1

DEAE

buffer C

jabuffer

C

0.10M KC1

Sehcl

0.10M KC1

0.15-0.20M KC1

buffer C

0. 1014 KC1

Superc

12 Fractions

13-14 V2

0.15-0.20M KC1

:)SO buffer C

0.10M KC1 Fractions

10-11

[image:3.612.84.269.65.369.2]

V4

FIG. 2. Flow diagram of theprocedures for purification of the factors forming the V2 and V4 DNA-protein complexes. HSV-1 infections in HeLacells, preparation of the nuclearextracts, and chromatographic purificationof theproteinsaredescribed in

Mate-rials and Methods.Column sizes used in these purifications

proce-duresare6 mlof P11 phosphocellulose, 3 ml ofheparin-agarose, 1

mlofDEAE-Sephacel,24 mlof Bio-Gel A-0.5M(0.8 by47cm), and 25 mlof fastprotein liquid chromatography Superose 12.

abilitytoformcomplexeswithprobeDNA ingelretardation

assays. Theactive fractionswere pooled.

Forheparin-agarosecolumnchromatography, active frac-tions eluted from the P11 column were pooled and diluted

with buffer CtoaKCI concentration of 0. 15 M and loadedon a preequilibrated heparin-agarose column (Sigma Chemical Co., St. Louis, Mo.). After the column was washed with 2 column volumes of buffer C-0.15 M KCI, a step gradient consistingof buffer C with a salt concentration from 0.2 to 0.5 MKCIwasdeveloped andfractionswere collected and

assayed by the binding gel retardation assay. Active frac-tions were pooledandprocessed.

For DEAE-Sephacel column chromatography, protein fractions pooled from the preceding column were diluted

with buffer Cto a KClconcentration of 0.1 M and loaded

ontoapreequilibratedDEAE-Sephacelcolumn(Pharmacia). Elution procedures werethesameasabove.

For Bio-Gel A-0.5M(25 ml; Bio-Rad Laboratories, Rich-mond, Calif.) orfast-protein liquid chromatography

Super-ose12(24 ml; Pharmacia)chromatography, proteinfractions

wereconcentratedtoless than 200 pJ,1% of the column bed volume, withaCentricon 30 concentration (Amicon Corp., Danvers, Mass.) prior to loading. After loading, proteins

wereeluted with buffer Ccontaining0.1MKClandfractions

were collected on a fraction collector and assayed for binding activity.

Gel electrophoresis of proteins. Electrophoretic separa-tions of denatured proteins were done in 9.25% polyacryl-amide gels cross-linked with N,N'-diallyltartardiamide con-taining sodiumdodecyl sulfate as described by Morse et al. (24). The gels were stained with silver for visualization (20) and mounted forphotography.

RESULTS

Experimental design. Preliminary experiments established that nuclear extracts of infected cells contained factors capable of binding to DNA fragments derived from the a sequence. Intheexperiments we reporthere, labeled DNA fragments were usedasprobes to monitor the purification of these factors, to determine the specificity of binding of the viral proteins recognizing DNA sequences contained in the terminal a sequence, and last, to identify the proteins involved in forming the complexes. As shown in Fig. 1, probe A contained the reiterated sequence DR4 and one copyofthe DR2 sequence. ProbesB, D,C,and Jcontained all or portions ofthe DR4,

Uc,

and DR1 domains of the a sequence. ProbesG, H, and E similarly covered the Ub and

DR1 domains of the a sequence. Last, probe F containing

unique sequences of the L component was used as the

representative of non-a sequence domains of the HSV-1

genome. Probes B and C were used for detecting and

monitoring the purification of viral proteins capable of

binding to the a sequence. These probes overlap; whereas

the C probe contains most if not all of the Pac2 sequence

required forcleavage recognition and the cleavage site, the B

probe contains only the cleavage recognition sequence. All

theprobes wereusedfordefiningthe specificity of binding.

Duringpreliminary experiments involvingDNA bandshift

assays, weidentified in the nuclearextracts of infectedand

mock-infected cells several factors capable of binding the

labeled DNA probes used in these studies. Of particular

interest were two factors designated V2 and V4 detected

only in extracts of infected cells. To differentiate between

these factors and to study their properties, we used the

concentrationand partial purificationscheme shown inFig.

2.

Demonstration of factors from infected cells capable of binding to the DNAprobe sequences. Figure 3 shows DNA

band shiftassaysutilizing probesB (panelsAand B) andC

(panels C and D) and both crudeandP11-fractionated (Fig.

2) nuclear extracts ofinfected andmock-infectedcells. The resultswere as follows.

(i) Both labeled DNAprobes formed multiple bands

con-taining DNA-protein complexes when reactedwith nuclear

extracts ofmock-infected and HSV-2(G)-infected cells. In

the experiments shown, the competitor DNA was

poly(dI-dC) poly(dI-dC), but substantially identical results were

obtained with other synthetic polymers

[poly(dI-dC);

data

not shown).

(ii)Severalof theDNA-protein complexes(e.g., the band

designated asHi in Fig. 3A and B) were detected in DNA

band shift assays withextractsof nuclei from mock-infected andfrom HSV-1- or HSV-2-infected cells. In contrast, the

DNA-protein complexes designated V2 and V4 were

de-tectedonlyin assays donewithextractsofinfected cells. In

general,thehost

protein-DNA complexes

tendedtoobscure

theinfectedcell-specific protein-DNA complexesformed

by

crude extracts ofinfected cells

(for example,

comparethe lanes marked M

[mock]

and G

[HSV-2]

shown between

panels with the lanes showing the reactions of fractionated

proteins in extracts of mock-infected

[panels

B and

D]

and

on November 10, 2019 by guest

http://jvi.asm.org/

1062 CHOU AND ROIZMAN

A

G:

P-fl

B _{M P-11} C _G

: P-l D M: P-1I

....

Lo

U) 'rufC2K0 U)OO()*n (_ :,a)0 r3 a)Unr) 'rC.<3il'r

C:>ICcDc3a oC.- 5OC M F G C OO O Oc a a °

-Hl1

V4

_Irb...

V2

14;1~~~~~~~~~~~~~~~~~~~~~~

CDU) C)U') CnU:S C U') Lo CD

OsCDDOC:3 0 F G O COO CD°

V4 X 4

*

i~~a-*.

i

Probc C

FIG. 3. Autoradiographic images of labeled DNA-protein complexeselectrophoretically separatedin nondenaturing gels. DNA probe fragments(Fig. 1)generated bydigestionof theplasmidswiththeappropriate enzymesweredephosphorylatedwithcalf intestinal alkaline phosphatase, purified frompolyacrylamide gels, and 5'end labeled with[.y-32P]ATP (>7,000Ci/mmol)byT4polynucleotide kinasetoan

activity of approximately 30,000cpm/ngofDNAfragment (17). (A and B) Reactionofprobe B DNAwith chromatographicfractions of nuclearextractprepared fromHSV-2(G)-infected cells (A) and mock-infected cells(B).(Cand D)Reaction ofprobeCDNAwith thesame

chromatographic fractions as in panels A and B. Lanes M, F, and G, Reactions of nuclear extracts of mock-, HSV-1(F)-, and HSV-2(G)-infected cells. Numbersatthe top of thegel indicatetheKCI concentration inbuffer C usedtoelute theproteinfractionsfrom column. V2andV4designatetheviralprotein-DNAcomplexes. Hi is discussed in thetext.

infected [panels A and C] cells). Moreover, the infected cell-specific protein-DNA complexesformed byextracts of HSV-2(G)-infected cellswere strongerandmore prominent

than those formed byextracts ofHSV-1(F)-infected cells. (iii) The properties and the specificity ofbinding of the factors responsible for the V2 and V4 DNA-protein

com-plexes aredealtwith in the text below. Of the hostfactors, the Hi factordefinedby P11 column chromatography (Fig. 3)wasshowninotherexperimentsdescribed later in thetext tobindstronglytoprobes B, J,and D butnottothe Cprobe. The designation of Hi assigned to the comigrating desig-nated bands inFig. 3A andB isprovisional; either the host DNA-binding proteinsin the mock-infectedcell extract are

different fromthose detected in the infected cellextractsor

theirpropertieshave beenalteredduringinfection inasmuch

astheir elutionprofiles were reproduciblydifferent.

Concentration andpartial purificationof V2and V4 factors from nuclear extractsof infected cells. The protocol for the purification of the proteins involved in the formation of V2 and V4DNA-protein complexes is shown in Fig. 2 and is

described inMaterials and Methods and in thelegendtoFig.

2. ThepurificationwasmonitoredbyDNAbandshiftassays

(Fig. 4) andbysilverstaining of proteins

electrophoretically

separated in denaturing 9.3%

polyacrylamide

gels (Fig. 5A

and B). In each step of the

purification procedure,

the fractionscontainingtheactivitydetected

by

DNA bandshift assayswere

pooled

and usedas

starting

materialin thenext step.

Asnotedearlierin the text,activity ofthefactorsin the V2

and V4 DNA-protein complexes were

barely

detectable in

crude nuclear extracts of HSV-1-infected cells but were more abundant in those of HSV-2(G)-infected cells. The factors forming the V2DNA-protein complex eluted in the

flowthrough fraction from P11 columnchromatography and

at aslightlylower salt concentration than V4from

heparin-agarose chromatography (Fig. 4). The Superose 12column fraction containingmost ofthe V2 activitymaterialyielded

on electrophoresis in a denaturing sodium dodecyl

sulfate-polyacrylamide gela

single major

silver-stained

polypeptide

band with anapparentmolecular

weight

of85,000for HSV-J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.612.112.517.72.430.2]

HSV-1 a SEQUENCE-SPECIFIC PROTEINS 1063

B. Heparin Agarose

I~~n a i~nn QCin

N

C%N

MFGC eaooa o oo

t

p

_ eu*

C. DEAE Sephacel

MFQ

an

a Co

MFG0 o o o o

. _-. -iw-- 1

0. Superose12

M F G P 3 4 5 B7

s1O12U14151S17*1S2l2;

8

W be

V2

a

_V2

AIAa--.__LAs

E.P11 F. HeparinAgarose

0.25 0O925

P11 un o C? ° P11 o

MFG' FG _W-gIci CD r-0 CD° M1: GNMf G o

4

G. OEAE Sephael

o an a cMs Go a o2o ocoa M F Go o c

o o o es

V4 4

Wv4

iqp~

~~~~~~W

3

IV4 swOr we

V4

a. L

FIG. 4. Autoradiographic images of labeledDNA-protein complexes electrophoretically separated innondenaturing gels. (AtoD)The

autoradiograms show the binding activityof theV2proteinfractionatedinsequentialorder by (A) P11 phosphocellulose,(B)heparin-agarose, (C) DEAE-Sephacel, and (D) Superose 12 columns.Thechromatographicprocedureswere asdescribed in the legendtoFig. 2.LanesM, F,

and Gcontainingreactionmixturesofnuclearextractsofmock-,HSV-1(F)-,andHSV-2(G)-infectedcells,respectively,withprobe CDNA

indicate the position of V2 protein-DNA complexes. Fractions collected from each column were identified by the KCI content (P11, heparin-agarose,andDEAE-Sephacel)orserially (Superose 12). (EtoH)Sameas panels A and DexceptthatprobeBDNAwasusedto

probeV4DNA-protein complexes.LanesM, F,and G, Reaction mixtures of nuclearextractsof mock-,HSV-2(F)-,andHSV-2(G)-infected cellswithprobeDNAs. ThelettersV2and V4markthe position ofthecorresponding bands;asnoted in thetext,thesebandsarefrequently

notseeninassaysof crude nuclearextracts.

1(F) (Fig. 5, right panel) or 82,000 for HSV-2(G) (data not shown) and avery faint band with an apparent molecular weight of60,000.Thefaintbandwasnotalwayspresent;it is

seen in lane 4 but not in lane 3 of Fig. 5 and may be a

degradation productof thehigher-molecular-weight protein. The material eluting from the Superose 12 column and forming the V4 DNA-protein complex formed two,

appar-ently equimolar, major silver-stained bands with apparent

molecular weights of >250,000 and 140,000 (Fig. 5, left panel),referredtoasproteinsAandB,respectively. Similar bands were obtained from extracts of HSV-2(G)-infected cells,andanalyses done with monoclonal antibodies reactive with infectedcell protein 1(ICP1) suggest that proteinAis ICP1 (data not shown). It should be stressed, as will be shownbelow,that atallstages ofpurificationof theproteins formingthe V4complex,thefractionscontainingtheactivity

A.P11

c MFG

V2*9 V2

H. Superose 12

MFG8 101112 .AMAIL A&Ak

mm

VOL.63, 1989

V4

on November 10, 2019 by guest

http://jvi.asm.org/

1064 CHOU AND ROIZMAN

HSV-1(F):V4 HSV-1(F):V2

KPG M F'4 3 2 1 F 1 2 3 4 ICP

4

4

11

S

17

17

* * _. 19

2425

[image:6.612.64.304.73.356.2]

.- II^

FIG. 5. Photographofasilver-stained denaturing (sodium

dode-cyl sulfate)9.3%polyacrylamide gelcontainingelectrophoretically separated proteins from various stages ofpurification ofproteins contained in V2 and V4DNA-protein complexes.LanesM, F,and Gwere loaded withapproximately20 jig eachof the crude nuclear

extracts ofmock-infected, HSV-1(F)-infected, or

HSV-2(G)-infec-ted cells, respectively. Lanes 1, 2, 3, and 4 in both left andright panelswereloaded withpooled activefractioncontainingeither V4 (left panel)orV2(right panel) activityandobtainedduringstepwise

purification of the proteins formingthe V2 and V4complexesfrom HSV-1(F)-infected cell extracts by chromatography on P11

phos-phocellulose (lanes 1), heparin-agarose (lanes 2), DEAE-Sephacel (lanes 3), and Superose 12 (lanes4), respectively. Thepurification protocol is shown in Fig. 2. Numbers to the right of the figure indicate the ICP numberdesignations of Honess and Roizman (12)

asamendedby Morseet al.(24).

yieldedtwobands in DNAretardation gels with probe B and threebands with probe C.

Biologic properties and identificationofprotein formingthe V2 DNA-protein complex. During studies of the biologic properties of the proteins forming theV2complex,wenoted

thataddition of the purified V2 (Fig. 5,right panel, lane 4)to

the labeled DNA probes in the presence of 10 mM Mg2+ resulted in the loss of the terminal 32p label from the probe DNAs (Fig. 6, right panel). One hypothesis that could account forthe apparentloss of thelabeled DNA probes is that addition of

Mg2'

activated a DNase. To test the possibilitythat theprotein containedin theV2DNA-protein complexistheviralDNase, weperformed band shiftassays

in the presence and absence of monoclonal antibody Ql

obtained from K. L. Powell of the Wellcome Research Laboratories and directed against the HSV-1 and HSV-2

enzymes. Earlier we showed that antibody specific to the DNA-binding proteins frequently binds the protein in the DNA-protein complexandfurther retards the mobility of the DNA(14). The resultsshown in Fig. 6,left panel, and other studies not shownindicated thefollowing.

(i) Only a singleDNA-protein complex was observed in all the assays done with partially purified fractions of the V2 factor (Fig. 5, right panel, lane 4). In each instance, the addition of the Ql monoclonalantibodyretarded the

migra-tion of the V2 complex, as would be expected from the

specific interaction of the monoclonal antibody with the protein factorcontained in the complex (14). A faint DNA-protein complex was also noted in the assays done with crude nuclear extracts ofHSV-2(G)-infected cells and the E probe. Themigration ofthis complex was alsoretarded by the antibody.

(ii) The formation of the V2 DNA-protein complex in assays with the probe F DNA, a 48-bp fragment derived from the promoter-regulatory domain of the a27 gene (13),

indicated that V2 factoris notspecificfor DNA sequences

containedin the asequence. The reliability of the band shift

assays is underscored by the observation that the F probe DNAdid react with theaH1proteinand thevirus-specifica trans inducing factor (aTIF) to produce the DNA complex

showninFig. 6. Theidentityof thecomplex was verified by

thedemonstrationthatantibody to the aTIF protein retarded

themigrationofthecomplex innondenaturing gels (datanot

shown), as was previously demonstrated (19).

Bindingspecificity of V4. Thepurified fraction which forms the V4 DNA-protein complex consists of two equimolar amounts of viral proteinscorrespondingin apparent

molec-ularweighttoICP1 and to a yet unidentifiedprotein(Fig. 5,

left panel). To define the specificity ofthe binding ofthese proteins to DNA, we tested alltheprobesgenerated inthis study for theirreactivityandabilitytoform theV4 protein-DNAcomplex with thepurifiedV4fraction shownin Fig.5, leftpanel, lane 4, and with crude nuclear extracts of

mock-infected cellsand cells infectedwithHSV-1(F). Theresults

indicate that probes D and C reacted with the purified

proteins formingthe V4complex toyieldthreebands. Bands

2and 3formedwith allprobes,except the bandsformedwith the J probe werebarely visible andthose formedwiththe L

andGprobeswerefainterthan thoseobtained with the other

probes. Band 1 was reproducibly present only in lanes

showingtheprotein complexesformed by probes D andC.

DISCUSSION

We identified several DNA-protein complexes that were

formed by nuclear extracts of both infected and mock-infected cells and two DNA-protein complexes designated V2and V4 that werespecific for nuclear extractsofinfected cells. Thefocus ofthestudies reported here was the

identi-fication of viral proteins which reacted with the a sequence

of HSV-1. It is convenient to consider the identity and

significance ofthe proteins formingthe various complexes

separately.

V4protein-DNA complex. The most purified preparation which formed the probe C DNA-V4 protein complex con-tained two viral proteins in approximately equimolar amounts. Protein A was identified as ICP1 by itsapparent molecularweight and its reactivity with a monoclonal anit-body (9). Protein B comigrated with a protein of a molecular weight predicted forICP7, but itsidentity is notknown. Of the three bands formed with probe C, the two designated bands 2 and 3(Fig. 7) were alsoformed with other probes, whereas thecomplex designated as band 1 formed only with

probes D and C, i.e., the probes that contain at least a

portionof the

Uc

sequence and DR1. We mayconcludefrom

these observations that either the specificity of binding

requires the interaction of both proteins or one of the two

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

Qi Oi Oi Oi

+ , t

t F OGV2V2M F V2WN

h IT I I

V4 .

V2

%

w.

0101 O Qi

+ + + I

AF GGV2V2M F GGVT2

_1

i

aTIF 1 ~2 3

Mg

3 -.... MW

4 5

_IM

E

240

bp

S

125bp

2~ 86

bp

q_p

56 bp

4

39bp

[image:7.612.98.521.76.370.2]

Probe:E D A F

Probe

C

with

V

2

Protein

FIG. 6. Autoradiographic images of labeled DNA electrophoretically separated in nondenaturing gels. Right panel: Images of purified V2 protein-labeled probe C DNA complexes formed in the presence or absence ofMg2+in thebinding reaction. Lane 1, Labeled probe without protein added; lane 2, probe DNAand 0.2 ,ul ofV2fraction containing noMg2". Lane 3, Same as lane 2 except that the binding buffer contained10mMMg2+.Lanes 4and 5,Sameaslanes2and 3, respectively, except that theconcentration of V2 protein was 0.5p.1perbinding reaction. The V2 preparation istheone shown in Fig. 5, right panel, lane 4. The probe C DNA and the V2fraction in a total of 20 ,u1were reacted at37°C for 1 h. The reaction was terminated by the addition of sodium dodecyl sulfate and proteinase K to final concentrations of 0.4% and100,ug/ml, respectively. The DNA was extracted with phenol, ethanol precipitated, dried, and subjected to electrophoresis on a 5% nondenaturinggel asdescribed intheResults. Left panel: Reactivity of V2 protein-DNA complexes with monoclonal antibodyQlspecific for HSV DNase. Lanes M, F, and G, Reactions of nuclear extracts of mock-infected, HSV-1(F)-infected, and HSV-2(G)-infected cells described above with various probe DNAs; lanes designated V2, reaction of labeled DNA probes identified on the bottom of the lanes with purified V2 protein shown in Fig. 5, right panel, lane 4. In lanes designated +Q1, the reaction mixture contained in addition to the stated ingredients1 p.lof monoclonal antibodytoHSV-1 DNase. The dot to the right of the band identifies the V2 protein-DNA complexes. Note thatprobeFcontains the cis sitefor the induction of a genes by the virion aTIF (19). With HSV-1(F)- or HSV-2(G)-infected cell extracts and the competitor DNAs, probeFDNAformscomplexes with a host protein designated as aH1(15) and with both aH1 and aTIF. The Ql antibody hasnoeffect on the formation ofaH1oraTIFbands (comparelane G, probe F, with lane G+ Ql, probe F). V2 forms a complex withprobe F, and this complex is retarded in its electrophoretic mobility by antibodyQl (rightmost two lanes). MW, Molecular weight.

proteins was copurified adventitiously and binds to DNA

nonspecifically.

ProteinsA (ICP1) andB have notbeen extensively

char-acterized. ICP1 isastructural protein(VP1); it is located in

the tegument and is not readily removed from virions

de-enveloped by treatment with nonionic detergents (16).

Fewer than 150 molecules of ICP1 per virion have been

estimated to bepresentbased on its then estimated

molec-ular weight of 260,000 to 270,000 (9). Since its translated

molecular weight is currently predicted tobe 336,000 (18),

the actual numberof molecules per virion might be

signifi-cantly smaller.

Independently derived data link ICP1 with encapsidated

DNA and more specifically with the release of viral DNA

from capsids (1). Thus, in cells infected with

HSV-1(HFEM)tsB7 at the nonpermissive temperature, the

cap-sidsaretransportedtothenuclear pore, but viral DNA isnot

released into the nucleus. The DNA is released upon shift

downtothepermissivetemperature. The mutation in

HSV-1(HFEM)tsB7 wasmapped by marker rescuetechniques to

thedomain oftheICP1gene. Itis conceivablethatICP1may

be involved in the shuttling ofthe DNA in and out ofthe

capsids, but atleast the packagingfunction remains to be

proved.

The genespecifying proteinB has notbeen identified. In

earlierstudies, proteinBwouldhavebeen obscuredbyother

virion proteinswith similarelectrophoretic mobilities(9).

V2 protein-DNA complex. The most purified V2 protein

preparation containedasinglepolypeptide.The

electropho-reticmobility ofthe protein, its enzymatic activity, and its

reactivitywith monoclonal antibodyare consistent with its

beingthevirally encodedDNase. Thebindingof the DNase

toDNA, although notunexpected, has notbeenpreviously

reported. Although the virally encoded DNase has strong

exonucleolytic

activity,

it has been

reported

to exhibit

non-sequence-specific endonucleolytic activity in the

pres-enceof

Mg2+

(10,11). Thebindingof the DNasetotheviral

DNAfragments doesnotappeartobe sequence

specific.

It

0 acHi

on November 10, 2019 by guest

http://jvi.asm.org/

1066 CHOU AND ROIZMAN

TABLE 1. Correlation ofsusceptibility to cleavage with the formation ofsequence-specific complexes

Sequence present

Probe asequencecomponent present Cleavage inampliconorvirus

cissite

A DR2 + (DR4)3 - - - NDa

B Uc + portionofDR1 - + - ND

C Portion ofUc + DR1 (portionofprobe D) - + + ND

D Uc+ DR1 - + + +(6)

E Ub+ DR1 + - + -(6)

F None(regulatory region of the ca27 gene) - - - ND

G Portion ofUb+ DR1(portion ofprobe E) - - + _b

H Ubsubfragment (portion ofprobe E) + - - ND

J PortionofU,, (portion of probe D) - - - ND

L Uc, sameasprobe D,but minus the BssHIIfragment - - -C

a ND, Not determined.

bThis is the smallestfragment containinganintact DR1 sequence. The DR1 sequencewastestedin the virus and in theampliconsystemwithnegativeresults

(6).

cTheeffect of the deletion of the BssHIIfragment fromUcwastested inthe contextofalargerfragment containing DR1,theUcminusthe BssHIIfragment,

and thereiterated DR4 and DR2 sequences andUbsequence.TheDNAfragmentwas notcleavedin theampliconsystem,indicatingthat the deleted sequence wasessential for cleavage (6).

should be noted, however, thatDNA binding andcleavage

by the DNasecould be sequence specific ifits activitywas

directedbya protein whichcoupled tothe DNaseandwas

bound to the viral DNA in a sequence-specific fashion. A

MFV4MFV4MFVY4MFV4MFV4 MFV4 MFV4MFV4 MFV4

w..t~~~~~~~~~~~~~

2

fti,4 | i s

a

\ t

~i'

w

0

.:#I

p

4-role for the viral DNase may also be deduced from the available data on thecleavage ofunit-lengthDNAmolecules fromconcatemers orcircles.

Specifically, (i) the Lcomponent terminus and the

junc-tion between the L and S components of the HSV-1 DNA

eachcontainone to several a sequences. The Scomponent

terminus containsonly one a sequence(23). Studies onthe

structureof the asequences attheterminiandthe

junction

of theLand Scomponentsledtotheconclusion that the site

of the cleavage of concatemers orcircles into linear

unit-length DNAmoleculesiswithintheDR1 sequence shared

by

two adjacent a sequences and results in single-base 3'

extensions in the ends of the linear molecules (23). The

endonuclease

responsible

for the

cleavage

of the DNA has

not been identified. Although the cleavage and

packaging

appear to be related (7), it is not inconceivable that viral DNase in connection with other site-directed proteins is

responsiblefor thiscleavage. The HSV DNase can act asan

endonuclease.

(ii) Akey question regarding

cleavage-packaging

of viral

DNA was the fate ofnewly synthesized molecules with a

singleasequence at the componentjunctions. Insertion ofa

singleasequence into virus resulted in oneof two events. In

someinstances,the asequence wasreduplicatedintandem,

yieldinguponcleavagetwoidentifiable endseach

containing

an asequence. When the a sequencewasnotamplified,the

cleavageeventoccurredin thedistal

(post-U,)

DR1and the

terminus containing solely the residualportion of DR1 but

not the remaining components ofthe a sequence was no

longer detectable andwaspresumedtobedegraded (28). If

the same sequence of events was to occur in

cleavage-D C J B

PROBE

FIG. 7. Autoradiographic images of the complexes formed by labeled DNA probes with crude extracts and purified fractions containing the proteins forming the V4complex. The complexes formedby each of theprobes shownonthebottom of thefigure with nuclearextracts of mock-infectedcells (M)or HSV-1(F)-infected

cells(F)orwithpurifiedV4

(V4)

areshown in the three

correspond-inglanes ofeach

triplet.

The V4 bands

unique

to

probes

Dand Care

identifiedasband1.Bands 2 and 3 appeartoform withmostprobes tested inthis study. The purified V4is that shown in Fig. 5, left

H G panel, lane4.Thecomplexeswereelectrophoretically separatedon

EHG nondenaturing gels. The bands formed withextractsof infectedand mock-infected cells and migrating slightly slower than band 1 correspondtothosedesignatedHi in Fig.3.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:8.612.68.562.87.224.2] [image:8.612.60.303.344.721.2]

packaging

of viral genomes from concatemers, it wouldbe

expected

that theDNAmolecule

lacking

anasequenceatits

terminuswould be

degraded exonucleolytically

untila

com-ponent

junction containing

acompetentasequence wouldbe

reached. The HSV DNasecan act as anexonuclease. Possible role of

proteins forming

the V4 complexin

cleav-age-packaging

of HSV-1 DNA.The

significant

aspectsofthe interaction of the

proteins

in the

purified preparation

with the DNA

probes

rest onthecorrelation betweenthe forma-tion of the

sequence-specific

band 1in

Fig.

7and the in vivo

ability

of the

probe

sequencesto act astargetsfor

cleavage-packaging

inviral DNAorin

amplicons (Table 1).

Previous studieshaveshown that the relevant

cis-acting

sites for bona fide

cleavage

areinthesequences

designated

Paclin

Ub

and Pac2 in

Uc

and,

in intact a sequences, the

intervening

DR1 sequence whichcontains the actual siteof

cleavage

ofviral

DNAfrom circles orconcatemersinto

unit-length

molecules

andfor

packaging

into

capsids

(6, 7, 28).

Theresults

reported

here indicatethatthe

complex exemplified by

band 1in

Fig.

7 isformed

only

with

probes

D andCwhichcontainthePac2 and

DR1

sequences but notwith

probe

E

containing

Pacl and theDR1sequence. The upperband

(band 1)

didnotform with DNA

fragments containing

only

DR1 or Pac2 (e.g.,

probes

B and

G).

The concordant dataare thatinsertion of

probe

D into

amplicons

or into viral DNA results in the

cleavage

of the DR1 of the

probe

sequence, whereas the

insertion of

probe

E or of the

appropriate

DNA

fragments

withadeletion identicaltothat in

probe

Ldoes notresultin

cleavage (6).

Theseobservations

predict

that the V4

proteins

formpart ofthe

complex recognizing

the

DR1-Uc

sequence

required

for the

cleavage-packaging

of the viralDNAsfrom

concatemers. Theidentification of the viralDNaseasoneof the

proteins binding

to ourDNA

fragments,

albeit

nonspe-cifically,

may be

fortuitous,

butatthistimewe cannot

reject

the

possibility

that

cleavage

of the DR1 sequence

by

an

endonuclease is

specifically

directed

by

the V4

protein

complexes and,

as noted

above,

that viral DNase has the

appropriate

attributestofulfill thisrole. ACKNOWLEDGMENTS

We thank K. L.Powell,Wellcome ResearchLaboratories,for the monoclonal antibody Ql to the viral DNase and J. Wesley, the

University

of

Chicago,

for use of the Superose 12 column for

purification

of

proteins

in this work.

These studies were aided by Public Health Service grants

CA08494,

CA19264, and CA47451 from the NationalCancer

Insti-tute and A1124009 from the National Institute for Allergy and Infectious Diseases andbygrant MV2X from the American Cancer

Society.

LITERATURE CITED

1. Batterson, W., D. Furlong, and B. Roizman. 1983. Molecular

genetics

ofherpes

simplex

virus. VIII. Further characterization

ofatemperature-sensitive mutant defective in release of viral

DNA and in other stages of the viral reproductive cycle. J. Virol. 45:397-407.

2. Chou, J,and B. Roizman. 1985.Isomerization ofherpes simplex

virus1 genome: identificationof thecis-actingand recombina-tional sites within the domain of the a sequence. Cell 41: 803-811.

3. Chou, J.,and B.Roizman.1985. The terminalasequence ofthe

herpes simplex

virus genome containsthepromoter ofagene located intherepeat sequences ofthe Lcomponent. J. Virol. 57:629-637.

4. Davison, A.J. 1984. Structure of the genome termini of

vari-cella-zostervirus.J. Gen. Virol. 65:1969-1977.

5. Davison,A.J.,andN. M. Wilkie.1981. Nucleotide sequencesof

thejoint between the L and S segments ofherpessimplex virus type 1 and 2. J. Gen. Virol. 55:315-331.

6. Deiss, L. P.,J. Chou, and N. Frenkel. 1986. Functional domains within the a sequence involved in the cleavage-packaging of herpes simplex virus DNA. J. Virol. 59:605-618.

7. Deiss, L. P., and N. Frenkel. 1986. The herpes simplex ampli-con:cleavage of concatemeric DNA is linkedtopackaging and involves the amplification of the terminally reiterated a se-quence.J. Virol. 57:933-941.

8. Dignam, J. D., R. M. Lebovitz, andR.Roeder. 1983. Accurate transcription initiation by RNA polymerase II in a soluble extractfrom mammalian nuclei. Nucleic Acids Res. 11:1475-1489.

9. Heine, J. W., R. W. Honess, E. Cassai, and B. Roizman. 1974. Proteins specified by herpes simplex virus. XII. The virion polypeptides of type 1 strains. J. Virol. 14:640-651.

10. Hoffman,P.J. 1981. Mechanism of degradation of duplexDNA by the DNase induced by herpes simplex virus. J. Virol. 38:1005-1014.

11. Hoffmann, P. J., and Y. C. Cheng. 1977.Thedeoxyribonuclease induced after infection of KB cells by herpes simplex virus type 1 ortype 2. J. Biol. Chem. 253:3557-3562.

12. Honess,R.W.,and B. Roizman.1984.Regulation of herpesvirus macromolecularsynthesis. I. Cascade regulation of the synthe-sis of three groups of viralproteins. J. Virol. 14:8-19. 13. Kristie,T.M.,and B. Roizman. 1984.Separation of sequences

definingbasalexpressionfrom thoseconferring a gene recogni-tion within theregulatory domains of herpes simplex virus 1 a genes. Proc. Natl. Acad. Sci. USA 81:4065-4069.

14. Kristie,T.M.,and B. Roizman.1986. a4, the major regulatory proteinofherpes simplexvirus type 1, isstably and specifically associated withpromoter-regulatory domains of a genes and of selected other viral genes. Proc. Natl. Acad. Sci. USA 83: 3218-3222.

15. Kristie,T.M.,and B.Roizman.1987. Host cellproteins bindto the cis-acting site required for virion-mediated induction of herpes simplexvirus 1 a genes. Proc. Natl. Acad. Sci. USA 84:71-75.

16. Lemaster, S., and B. Roizman. 1980. Herpes simplex virus phosphoproteins. II. Characterization of the virion protein kinase and of thepolypeptidesphosphorylated in the virion. J. Virol. 35:798-811.

17. Maniatis, T.,E. F.Fritsch,andJ. Sambrook. 1982. Molecular cloning, alaboratorymanual. ColdSpring Harbor Laboratory, ColdSpring Harbor,N.Y.

18. McGeoch, D. J., M. A. Dalrymple, A. J. Davison, A. Dolan, M. C.Frame,D.McNab,L.J. Perry, J.E.Scott,and P.Taylor. 1988. ThecompleteDNAsequence of thelong unique region in the genome of herpes simplex virus type 1. J. Gen. Virol. 69:1531-1574.

19. McKnight, J. L. C., T. M. Kristie, and B. Roizman. 1987. Binding of the virion protein mediating a gene induction in herpes simplex virus 1-infected cells to its cis site requires cellularproteins. Proc.Natl.Acad. Sci. USA 84:7061-7065. 20. Merril, C. R., D. Goldman, S. A. Sedman,and M. H. Ebert.

1981. Ultrasensitive stain for proteinsin polyacrylamide gels showsregionalvariation incerebrospinalfluidproteins.Science 211:1437-1438.

21. Mocarski,E.S.,L. P.Deiss,and N. Frenkel. 1985.The nucleo-tide sequence and structural features ofanovelUs-ajunction present in adefective herpessimplex virus genome. J. Virol. 55:140-146.

22. Mocarski,E. S.,L. E. Post, and B. Roizman.1980. Molecular engineeringoftheherpes simplexvirus genome: insertion ofa secondL-Sjunctioninto the genomecausesadditionalgenome inversions. Cell22:243-255.

23. Mocarski,E.S.,and B. Roizman.1982.Thestructureandrole of theherpes simplexvirus DNAtermini ininversion, circulariza-tion,andgenerationof virionDNA. Cell31:89-97.

24. Morse,L.S.,L.Pereira,B.Roizman,and P. A. Schaffer.1978. Anatomyofherpes simplexvirus DNA. XI. Mappingof viral genes by analysis ofpolypeptides and functions specified by HSV-1 x HSV-2 recombinants. J. Virol. 26:389-410.

on November 10, 2019 by guest

http://jvi.asm.org/

1068 CHOU AND ROIZMAN

25. Poffenberger, K. L., and B. Roizman. 1985. A noninverting

genomeofaviable herpes simplex virus 1: presence of

head-to-tail linkages in packaged genomes and requirements for

circularizationafterinfection.J. Virol. 53:587-595.

26. Roizman, B. 1979. The structureand isomerization ofherpes simplex virusgenomes.Cell16:481-494.

27. Strauss, F., and A. Varshavsky. 1984. A protein binds to a

satellite DNA repeat at three specific sites that would be broughtinto mutualproximity by DNA folding in the

nucleo-some.Cell37:889-901.

28. Varmuza, S. L., and J. R. Smiley.1985.Signals for site-specific cleavageof herpessimplex virus DNA: maturation involvestwo separatecleavageevents at sitesdistaltotherecognition site. Cell41:792-802.

29. Vlazny,D. A., A. D.Kwong, and N. Frenkel. 1982.Sitespecific cleavage/packagingofherpessimplex virusDNAandthe selec-tive maturation of nucleocapsids containing full length viral DNA. Proc.Natl. Acad. Sci. USA 79:1423-1427.

J.VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/